Br. J. Pharmacol. (1991), 104, 1045-1049

Selective antagonism of capsaicin by capsazepine: evidence for a

spinal receptor site in capsaicin-induced antinociception

A.H. Dickenson & l*A. Dray

Department of Pharmacology, University College, London and *Sandoz Institute for Medical Research, Gower Place,London WC1E 6BN

1 Capsazepine has recently been described as a competitive capsaicin antagonist. We have used thiscompound to test the hypotheses that the in vitro and in vivo effects of capsaicin are due to interactionswith a specific receptor.t2 In an in vitro preparation of the neonatal rat spinal cord with functionally connected tail, the activa-tion of nociceptive afferent fibres by the application of capsaicin, bradykinin or noxious heat (48'C) to thetail could be measured by recording a depolarizing response from a spinal ventral root. Application ofcapsaicin or substance P to the spinal cord also evoked a depolarizing response which was recorded in a

ventral root.3 When capsazepine (50nM-20pM) was administered to the tail or spinal cord it did not evoke any

measurable response. However on the tail, capsazepine reversibly antagonized (IC50 = 254 + 28 nM) theresponses to capsaicin but not to heat or bradykinin administered to the same site. Similarly capsazepineadministration to the spinal cord antagonized the responses evoked by capsaicin (IC50 = 230 + 20nM)applied to the cord but not responses evoked by substance P on the cord or by noxious heat and capsa-

icin on the tail.4 In halothane anaesthetized rats, C-fibre responses evoked by transcutaneous electrical stimulation ofthe receptive field were recorded from single wide dynamic range neurones located in the spinal dorsalhorn. C-fibre evoked discharges were consistently reduced by the systemic administration of capsaicin(20Oumol kg- 1, s.c.) and this action of capsaicin was antagonized by capsazepine (100,umol kg- 1) adminis-tered by the same route. In addition the systemic effect of capsaicin was antagonized by a spinal intrathe-cal administration of capsazepine (5-50 nmol).5 Intradermal injections of capsaicin, localized to the peripheral receptive field, usually one toe of theipsilateral hind-paw, produced a transient increase in C-fibre-evoked activity followed by a prolongedperiod of localized insensitivity to transcutaneous C-fibre stimulation. These effects of capsaicin were

significantly reduced by the concommitant administration of capsazepine to the same site.6 These data demonstrate that capsazepine is a selective antagonist of capsaicin on nociceptive neurones

in vitro and in vivo and suggest that the effects of capsaicin were mediated by activation of a specificreceptor. Since the antinociceptive effect produced by systemically administered capsaicin was antagonisedby spinal intrathecal capsazepine this further supports the hypothesis that capsaicin exerts its anti-nociceptive effect by acting on specific receptors localized to sensory nerve fibres in the spinal cord.

Keywords: Capsaicin; capsazepine; peripheral nociceptors; spinal cord; antagonist

Introduction

Capsaicin (8-methyl N-vanillyl-6-nonenamide) produces a

number of effects by selective actions on mammalian poly-modal nociceptive neurones and warm thermoceptors(Fitzgerald, 1983; Szolcsanyi, 1985; 1990; Buck & Burks,1986; Bevan et al., 1987). In man, local applications of capsa-icin produce a sensation of burning-pain due to the activationof primary afferent C- and probably Ab-neurones. However,with repeated exposure to capsaicin both the C-fibre activa-tion and the algesic effect desensitize (Petsche et al., 1983;Marsh et al., 1987). On the other hand acute systemic admin-istration of capsaicin produces antinociception which has beenshown by several conventional tests (Hayes & Tyers, 1980;Hayes et al., 1984; Campbell et al., 1989). The basis for thereversible antinociceptive effect of capsaicin is unclear thoughrecent studies (Dickenson et al., 1990a,b) have indicated thatsensory nerve terminals in the spinal cord are important.

It has been suggested that the effects of capsaicin on mam-

malian sensory neurones are mediated through an interactionwith a specific membrane receptor. However, the evidence forthis has been rather indirect. Thus a capsaicin recognition sitehas been proposed from structure-activity studies (Szolcsanyi& Jancso-Gabor, 1975; 1976; Hayes et al., 1984) and by the

1 Author for correspondence.

use of a capsaicin-like photoaflinity probe (James et al., 1988).Somewhat more directly, capsaicin has been shown to dis-place the binding of [3H]-resiniferatoxin (Szallasi & Blum-berg, 1990), a highly potent, naturally occurring capsaicinanalogue (Szallasi & Blumberg, 1989; Winter et al., 1990) witha similar mechanism of action (Winter et al., 1990). Recentlyhowever, a selective capsaicin antagonist, capsazepine, hasbeen described (Bevan et al., 1991; Dray et al., 1991) whichhas provided direct pharmacological evidence for a capsaicinreceptor localized on the mammalian sensory neural mem-brane.

In the present experiments we have used capsazepine toindicate whether the capsaicin-mediated activation of nocicep-tors in vitro or in vivo is a consequence of specific receptorinteractions. Furthermore we have sought additional evidencefor sensory elements in the spinal cord as primary targets forthe acute analgesic action of systemic capsaicin. Parts of thisstudy have been published as abstracts (Dickenson et al.,1991; Dray et al., 1991).

Methods

The in vitro preparation

The intact spinal cord and the functionally connected tailwere removed from 1-2 day old rats following decapitation

.Iv,'-. Macmillan Press Ltd, 1991

1046 A.H. DICKENSON & A. DRAY

and prepared by the method described by Otsuka & Yanagi-sawa (1988). The skin was carefully removed from the distalfour fifths of the tail. This procedure was thought to exposecutaneous fibres and their endings to facilitate activation ofnociceptors by capsaicin and other algesic agents (Dray et al.,1990a,b). Damage to the underlying tissue was avoided as thisseverely compromised responsiveness to peripheral stimuli.We cannot be certain however that our tests were made in anenvironment entirely free of tissue damage. Histology was notroutinely performed. Indeed electron microscopy would havebeen necessary to confirm the structural integrity of fine affer-ent nerve endings and quantification of this would have pre-sented considerable difficulty. However bradykinin, which wasused routinely as an algesic chemical stimulus, is not knownto activate axons of nociceptors. It is more likely to stimulatenociceptors via the signal transducing elements localized tothe terminations of primary afferent fibres. The efficacy ofbradykinin thus suggested that nociceptors were preserved inour viable preparations. This would strongly indicate minimalnerve fibre damage. In addition the effects of our peripheralstimuli were robust and reproducible over many hours. Thisalso would be unlikely in the face of significant tissue damage.The preparation was placed in a chamber such that the

cord and tail could be separately superfused (2-4mlmin-')with a physiological salt solution (composition mM: NaCl138.6, KCI 3.35, CaCl2 1.26, MgCl2 1.16, NaHCO3 21.0,NaHPO4 0.58, glucose 10) at 240C and gassed with 95%02/5% CO2. Peripheral nociceptive fibres were activated bysuperfusion of the tail with noxious chemicals (capsaicin,bradykinin) and by superfusate heated to 480C. Each stimuluswas applied for 10s with an intervening period of 15 minbetween stimuli. Bradykinin doses were separated by at least40-60 min to avoid tachyphylaxis.The activation of peripheral fibres was assessed by measur-

ing the depolarization produced in a spinal ventral root(L3-L5). The ventral root depolarization was recorded d.c.(with respect to the spinal cord which was earthed) with a lowimpedence glass pipette. This was placed in an electrolyte-filled well which contained the selected ventral root. Thesignals were amplified by conventional means and displayedsimultaneously on an oscilloscope and on a rectilinear chartrecorder.

Reproducible ventral root responses were obtained to per-ipheral stimuli including noxious heat and submaximal con-centrations of capsaicin and bradykinin. Following thiscapsazepine was administered in the tail superfusate for15min prior to retesting capsaicin and other stimuli. Theantagonist potency (IC50) of capsazepine was estimated bycumulatively increasing the concentration to produce anincremental reduction of the response to capsaicin. Three ormore submaximal concentrations of capsazepine were used ineach experiment to determine the IC50 concentration. Ventralroot responses were also used to measure the activation ofafferent fibres and postsynaptic elements in the spinal cordby spinal administrations of capsaicin and substance P respec-tively. The effect of capsazepine administered to the spinalcord was determined as described above against responsesevoked by administration of substance P and capsaicin to thespinal cord and to responses evoked by noxious heat and cap-saicin administration to the tail.

In vivo preparation

Single unit recordings were made from dorsal horn neuronesin the intact halothane-anaesthetized rat (Dickenson & Sulli-van, 1987). Following induction of anaesthesia (3% halothanein 02/N20) a tracheal cannula was inserted and a laminec-tomy was made to expose the spinal L,-L3 area. The animalwas then mounted in a frame so that the vertebrae rostral andcaudal to the recording site were securely clamped. Anaes-thesia was then maintained by 1-1.5% halothane. A tungsten-glass microelectrode was inserted into the dorsal horn andextracellular recordings were made, by conventional tech-

niques, from single neurones located in the superficial anodeep laminae. All cells responded to innocuous peripheralstimuli (touch/brush/pressure) and to noxious thermal andmechanical (pinch) stimuli. Cells were characterized by theirresponses to the natural peripheral stimuli mentioned aboveand then two fine needles were placed in the center of thehindpaw receptive field for transcutaneous stimulation. BothAfl and C-fibre responses, based on latency and thresholdcould always be elicited in the neurones, and on occasion Abfibre responses were observed. Following at least two stablecontrol responses (3 times threshold for the C-fibre and thenAfl-fibre responses), capsaicin or the vehicle (10% Tween/10%ethanol (10%); physiological saline) were administered at thefollowing sites: (1) subcutaneously into the scruff of the neck,(2) intrathecally in a volume of 50,ul onto the exposedsurface of the spinal cord and (3) locally, in a 10ul volumeinto a discrete region of the receptive field. Administration atthese sites was repeated with the co-administration of capsaze-pine or with capsazepine alone. The effect of intrathecal cap-sazepine was also tested against the effect of systemiccapsaicin. Responses to transcutaneous stimulation were mea-sured at 5min post injection and then at 10min intervals forbetween 45 and 180min. Responses were quantified by use ofpost stimulus histograms, analysed and built with a CED1401 interface and software (Dickenson & Sullivan, 1990).Only one cell was used per animal for each systemic drugstudy. Data were statistically evaluated by the unpaired t test(95% confidence limits).The following drugs were used: capsaicin (Sigma), cap-

sazepine (2-[2-4-chlorphenyl)ethylamino-thiocarbonyl]-7,8-dihydroxy-2,3,4,5-tetrahydro- 1H-2-benzazepine; bradykinin(Bachem, CRB).

Results

In vitro study

As in previous studies (Dickenson et al., 1990;. Dray et al.,1990a,b), a brief administration of capsaicin at a submaximalconcentration (usually 0.5-0.7 pM), in the tail superfusate,evoked a short lived (45-90 s) depolarizing response in aventral root. These responses could be reproduced overseveral hours. In addition stable and reproducible responsescould be evoked by brief noxious heat stimulation of the tail.In a number of experiments responses evoked by bradykinin(100-350nM) were used to test further the selectivity of theantagonist action of capsazepine.Continuous superfusion of the tail with capsazepine (50nM-

20pM) for 15min preceding and during the administration ofcapsaicin or other stimuli did not itself evoke any response.However capsazepine consistently and reversibly (within 30-60min) attenuated the effect of capsaicin in a concentration-related manner (Figure 1). The capsazepine IC50concentration was 254 + 28 nM (n = 8). However, even at thehighest concentration tested (20pM), capsazepine did not affectthe responses evoked by heat (n = 8) or bradykinin (n = 5)(Figure 1).

In a separate series of experiments, brief administration(10s) of a submaximal concentration of capsaicin (200nM), inthe spinal cord superfusate, evoked reproducible depolarizingresponses following the activation of nociceptive afferent nerveterminals. A selective effect of capsaicin on afferent terminalshad been indicated from previous studies showing (a)capsaicin-evoked release of neuropeptide specifically locatedin fine afferent fibres (Theriault et al., 1976; Dickenson et al.,1990b), (b) lack of effect of capsaicin on spinal neurones fol-lowing neurotoxic doses (Yaksh et al., 1979) and (c) lack of adirect effect of capsaicin on spinal dorsal horn neurones mea-sured electrophysiologically (Urban & Dray, 1990). Responseswere also evoked by administration of a submaximal concen-tration of substance P (25 nM), presumably through activationof receptors on postsynaptic spinal elements (Akagi et al.,

CAPSAICIN ANTAGONISM BY CAPSAZEPINE 1047

a

Caps Heat Caps BK0.3 fLM 0.6 p.M 0.3 p.M

CaCa

Nk'xPaps b.6 p.M Heat Caps 0.6 RM Heat3psazepine 50 niM Capsazepine 150 nM

Caps Heat BKCapsazepine 300 nM

AF~(~ ~'\. Wash 60 min

Caps Heat

10.2 mVIs

b

Heat Caps 0.2 FM SP 0.2 FxMS.cord S.cord

VA.EALHeat Caps SP

Capsazepine 500 nM

Wash 90 min

Spinal cord

change the responses evoked by spinal administration of sub-stance P. In addition, responses evoked by peripheral noxiousheat or capsaicin applications were unaffected (Figure 2).However, the responses evoked by capsaicin administration tothe spinal cord were consistently attenuated in aconcentration-related manner (IC50 = 230 + 20nm, n = 5). Ineach case recovery from the effect of capsazepine occurredwithin 30-60 min.

In vivo study

Administration into the receptivefield Eight dorsal horn neu-rones were studied. All had receptive fields in two or moretoes. Each neurone served as its own control. Thus the effectof capsaicin alone (200pmol) and capsaicin plus capsazepine(1 nmol) was tested on each cell. Each cell responded toC-fibre stimulation (3 x threshold) with a mean response of20 + 6 spikes per stimulus. After capsaicin, none of the cellsresponded to C-fibre stimulation for at least 1 h afterwards.Each cell was activated by All-fibre stimulation but theseresponses were unchanged by capsaicin. After the co-administration of capsazepine with capsaicin the effect of cap-saicin was completely abolished. Thus the mean C-fibreevoked responses of the population was 17 + 5 spikes perstimulus and this was not significantly different from valuesobtained following vehicle injections. The same concentrationof capsazepine also prevented the activation of the neuronesby capsaicin (Figure 2). Capsazepine alone (n = 4), had noeffect on cell firing or on C-fibre evoked responses.

Systemic capsazepine versus systemic capsaicin Capsaicin(20 ,umol kg- ) injected subcutaneously into the scruff of theneck at a dose found to produce antinociception both inbehavioural and electrophysiological studies (Campbell et al.,1989; Dickenson et al., 1990a), produced a maximal 66% inhi-bition of the C-fibre evoked responses of dorsal horn cells at40min post injection (n = 7). In three other animals theadministration of capsaicin was immediately preceded by a s.c.injection of lOOpmol kg-1 capsazepine into the same oranother s.c. site. In each of these studies the inhibitory effect of

xf t~~~~~~nHeat Caps SP

Figure 1 (a) Concentration-related antagonism of capsaicin (Caps)by capsazepine in the spinal cord/tail in vitro. The top traces showventral root responses produced by application of capsaicin (0.3 and0.6pM), heat (48°C) and bradykinin (BK, 0.3,pM) to the tail. Thesecond and third row of traces show concentration-related reductionof the response to capsaicin (0.6,pM) by capsazepine (50, 150 and300nM) but not the responses to heat or bradykinin (0.3,UM). Theapplication of capsazepine was started 15min before capsaicin was

tested and the concentration was increased cumulatively. Theresponse to capsaicin recovered following a 60min wash of the tissue(bottom traces). The calibration bars of 60s and 0.2mV are indicatedin (a) and (b). (b) Selective antagonism of capsaicin on the spinal cordby capsazepine. The top traces show control responses to heat admin-istered to the tail and to capsaicin (0.2pM) and substance P (SP,0.2 pM) administered in the spinal cord superfusate. During the contin-uous administration of capsazepine to the spinal cord (500 nM), begun15min before the agonists were retested, the effect of capsaicin was

selectively reduced. The bottom traces show complete recovery of cap-saicin responses after washing the tissue for 90 min.

1985). Such responses, together with the ventral rootresponses evoked by peripheral application of heat or capsa-icin, served as controls to check for selectivity of the effect ofcapsazepine.As before, prolonged superfusions of the spinal cord with

capsazepine (50nM-5pM) did not produce any response or

200'

I

I0wea) 4nr j._

Q)

0

Capsaicin 200 pmol+ capsazepine 1 nmolinto receptive field

Capsaicin 200 pmolinto receptive field

Time (s

Figure 2 Example of a ratemeter recording of a dorsal horn nocicep-tive neurone in response to transcutaneous electrical stimulation ofthe receptive field (two toes on the ipsilateral hindpaw). The combinedA/i and C-fibre evoked response is plotted as spikes s- (vertical axis)against time (s). The combined intradermal injection of capsaicin(200 pmol) and capsazepine (1 nmol) produced minimal C-fibre activa-tion and little evoked effect on the neurone (injected into toe 1),whereas capsaicin alone (injected into toe 2) activated the cell andthen abolished the C-fibre evoked response. The residual activity seenin the post-stimulus histogram after capsaicin was due entirely to theA/i-fibre evoked activity.

0 500 1000 1500 2000 2500

4

.. A

i .

1048 A.H. DICKENSON & A. DRAY

0U.

120 -

100 -

80 -

60 -

40 -

20 --20

120 -

100 -

o 80

0C.)

o 60

40

0 20 40 60 80

Time (min)Figure 3 The C-fibre-evoked responses of dorsal horn neurones,expressed as percentage (with s.e.mean shown by vertical bars) of thepre-injection control responses, were inhibited by s.c. capsaicin(20umol kg- 1, n = 7, 0). Co-administration of capsazepine(lOOumol kg- 1, s.c., n = 3, *) together with s.c. capsaicin produced asignificant reduction (P < 0.05), at the 30, 40, 50 and 60min timepoints. The control readings prior to the application of the agents attime 0 are shown.

capsaicin was now transient and at 30 min post injectionC-fibre evoked activity was depressed by only 13% (Figure 3).The response of cells to Af? fibre stimulation was unaffectedby capsaicin, capsazepine or by the combination of these sub-stances.

Intrathecal capsazepine and systemic capsaicin Neuronalactivity was recorded for at least 60 min following the intra-thecal administrations of capsazepine. No significant changein spontaneous firing of dorsal horn cells was observed and noeffects on responses evoked by C- or Afl-fibre stimulationwere produced by 5 nmol, (4 cells) or 50 nmol (3 cells) capsaze-pine.

In control animals (n = 7), s.c. capsaicin (20pmol kg-1) pro-duced a gradual reduction of the discharges evoked by C-fibrestimulation. This effect plateaued some 40 min after the injec-tion and reached a maximum of 66% inhibition. The Afl-fibreevoked responses of the same cells were maximally reduced byonly 25% at 30 min after the adminstration of capsaicin. Intra-thecal capsazepine (50 nmol; n = 5) significantly (P < 0.05)reduced (at the 40, 50 and 60 min time points) the inhibitoryeffect of systemic capsaicin (Figure 4). Following capsazepinethe responses of these cells to Afl fibre input was maximallyreduced by capsaicin by only 8%. When the C-fibre evokedresponses were analysed, taking into account the firing fre-quency and duration of the response, the intrathecalcapsazepine-induced reduction of capsaicin (by 58%, P < 0.01,Figure 4) was highly significant. Also the antagonism of capsa-icin by capsazepine appeared to be dose-related since the inhi-bition of C-fibre responses produced by capsaicin(20pmol kg-', s.c.) was unaffected by a lower dose (5nmol,i.t.) of capsazepine (n = 4).

Discussion

Indirect evidence obtained from a number of previous studieshas supported the suggestion that the unique effects of capsa-icin on mammalian sensory neurones was due to an inter-action with a specific membrane receptor (Szolcsanyi &Jancso-Gabor, 1975; 1976; Hayes et al., 1984; James et al.,1988; Szallasi & Blumberg, 1989; 1990). This hypothesis nowreceives direct support following the discovery of capsazepine,a capsaicin analogue with selective antagonist properties.Capsazepine antagonized a number of the effects of capsaicinin vitro, including the evoked whole cell membrane currents

-20 0 20 40 60 80Time (min)

Figure 4 The inhibition of the C-fibre responses of dorsal horn noci-ceptive neurones produced by s.c. capsaicin (20.umol kg- ', s.c., n = 7,0) was attenuated by intrathecal capsazepine (50nmol) applied justprior to the capsaicin (n = 5, 0). The capsaicin response was signifi-cantly reduced (P < 0.01) at the 40, 50 and 60 min time points. Appli-catoin of the drugs was at time zero.

and the flux of a number of cations (Bevan et al., 1991). Theantagonistic effect was both selective and competitive.The present in vitro experiments further extend these obser-

vations. We have thus shown, as in previous studies(Yanagisawa & Otsuka, 1984; Dray et al., 1990a,b) thatphysiological stimulation of peripheral nociceptive fibres bycapsaicin and a number of other noxious stimuli, can evokereproducible responses in the spinal cord, maintained in vitrowith the functionally connected tail. Capsazepine, at concen-trations which itself produced no discernible effect on periph-eral nerves or on the excitability of the spinal cord,consistently and reversibly antagonized the effects of capsaicinwhen both substances were administered to the same periph-eral or central elements of sensory neurones. Capsazepine didnot reduce the responses evoked by capsaicin when each sub-stance was administered to different regions (i.e. spinal vs.peripheral) of sensory neurones.

Although the present experiments did not test whether cap-sazepine was a competitive antagonist, they clearly demon-strated that capsazepine was selective, since responsesproduced by noxious heat, bradykinin or substance P admin-istered to the periphery or spinal cord respectively, were unaf-fected. The potency of capsazepine, under the presentconditions, was similar to that measured in dorsal root gang-lion neurones or vagal afferent fibres maintained in vitro(Bevan et al., 1991) where competitive interactions were alsodemonstrated. This supports the possibility that the capsaicinreceptive site is similar in a variety of in vitro preparations ofsensory neurones.

In addition we have tested capsazepine against a number ofeffects mediated by capsaicin in vivo: specifically against theeffects produced by local and systemic capsaicin adminis-trations. Thus capsazepine antagonized the initial capsaicin-induced C-fibre activation and the subsequentcapsaicin-induced inactivation of C-fibres following theadministration of capsaicin into the peripheral receptive field,(usually a single toe) of dorsal horn neurones. These effectswere clearly selective and well localized as a subsequentadministration of capsaicin into another part of the receptivefield (another toe), evoked both an initial activation and asubsequent blockade of C-fibre evoked activity. In keepingwith our previous observations (Dickenson et al., 1990a,b),systemic capsaicin also reduced C-fibre evoked responses ofdorsal horn nociceptive neurones. This antinociceptive effectof capsaicin was also reduced by the concomitant adminis-tration of capsazepine by the same route. Finally and signifi-cantly, spinal intrathecal administration of capsazepineprevented the reduction of the C-fibre evoked input produced

0

lll.

CAPSAICIN ANTAGONISM BY CAPSAZEPINE 1049

by systemic capsaicin. This was considered to be due to anantagonism of a capsaicin-induced effect exerted at the level ofthe spinal cord since (a) capsazepine itself did not produce aresponse or alter nerve excitability but acted only as a selec-tive capsaicin antagonist and (b) the amount of capsazepineadministered i.t. was insufficient to antagonize the systemiceffect of capsaicin should diffusion out of the spinal cord haveoccurred.

In summary, the present experiments have demonstratedthat capsazepine is a selective antagonist against a number ofcapsaicin-induced effects on nociceptive neurones in vitro andin vivo. In addition, the effectiveness of capsazepine followingspinal intrathecal administration further supports the likeli-hood that the systemic antinociceptive effect of capsaicin isproduced by an action within the spinal cord as we have pre-viously suggested (Dickenson et al., 1990a,b).

References

AKAGI, H., KONISHI, S., OTSUKA, M. & YANAGISAWA, M. (1985). Therole of substance P as a neurotransmitter in the reflexes of slowtime courses in the neonatal rat spinal cord. Br. J. Pharmacol., 84,663-673.

BEVAN, S.J., JAMES, I.F., RANG, H.P., SHAH, K. & YEATS, J.C. (1991).The development of a capsaicin antagonist for the sensoryneurone excitant, capsaicin. Br. J. Pharmacol., 102, 77P.

BEVAN, S.J., JAMES, I.F., RANG, H.P., WINTER, J. & WOOD, J.N. (1987).The mechanism of action of capsaicin - a sensory neurotoxin. InNeurotoxins and their Pharmacological Implications. ed. Jenner, P.pp. 261-277. New York: Raven Press.

BUCK, S.H. & BURKS, T.F. (1986). The neuropharmacology of capsa-icin: review of some recent observations. Pharmacol. Rev., 38, 179-226.

CAMPBELL, E.A., DRAY, A. & PERKINS, M.N. (1989). Comparison ofcapsaicin and olvanil as antinociceptive agents in vivo and invitro. Br. J. Pharmacol., 98, 907P.

DICKENSON, A., ASHWOOD, M., SULLIVAN, A.F., JAMES, I. & DRAY,A. (1990a). Antinociception produced by capsaicin: spinal or per-ipheral mechanism? Eur. J. Pharmacol., 187, 225-233.

DICKENSON, A.H., DRAY, A., HUGHES, G.A. & WALPOLE, C.S.J. (1991).The selective antagonist capsazepine inhibits capsaicin inducedantinociception: electrophysiological studies in rodents, Br. J.Pharmacol., 102, 79P.

DICKENSON, A., HUGHES, C., RUEFF, A. & DRAY, A. (1990b). A spinalmechanism is involved in the antinociception produced by thecapsaicin analogue NE 19550 (olvanil). Pain, 43, 353-362.

DICKENSON, A.H. & SULLIVAN, A.F. (1987). Subcutaneous formalininduced activity of dorsal horn neurons in the rat: differentialresponses to intrathecal opiate administration pre or post forma-lin. Pain, 30, 349-360.

DICKENSON, A.H. & SULLIVAN, A.F. (1990). Differential effects ofexcitatory amino acid receptor antagonists on dorsal horn noci-ceptive neurons in the rat. Brain Res., 506, 31-39.

DRAY, A., BETTANEY, J. & FORSTER, P. (1990a). Actions of capsaicinon peripheral nociceptors of the neonatal rat spinal cord-tail invitro: dependence of extracellular ions and independence of secondmessengers. Br. J. Pharmacol., 101, 727-733.

DRAY, A., BETTANEY, J. & FORSTER, P. (1990b). Resiniferatoxin, apotent capsaicin-like stimulator of peripheral nociceptors in theneonatal rat tail in vitro. Br. J. Pharmacol., 99, 323-326.

DRAY, A., CAMPBELL, E.A., HUGHES, G.A., PATEL, I.A., PERKINS,M.N., RANG, H.P., RUEFF, A., SENO, N., URBAN, L. & WALPOLE,C.S.J. (1991). Antagonism of capsaicin-induced activation of C-fibres by a selective capsaicin antagonist capsazepine. Br. J. Phar-macol., 102, 78P.

FITZGERALD, M. (1983). Capsaicin and sensory neurones - a review.Pain, 15, 109-130.

HAYES, A.G., OXFORD, A., RENOLDS, M., SKINGLE, A.H., SMITH, C. &TYERS, M.B. (1984). The effects of a series of capsaicin analogueson nociception and body temperature. Life Sci., 34, 1241-1248.

HAYES, A.G. & TYERS, M.B. (1980). Effects of capsaicin on nociceptive

heat, pressure and chemical thresholds and on substance P levelsin the rat. Brain Res., 189, 561-564.

JAMES, I.F., WALPOLE, S.J., HIXON, J., WOOD, J.N. & WRIGGLES-WORTH, R. (1988). Long lasting agonist activity produced by acapsaicin-like photoaffinity probe. Mol. Pharmacol., 33, 643-649.

MARSH, S.J., STANSFELD, C.E., BROWN, D.A., DAVEY, R. &McCARTHY, D. (1987). The mechanism of action of capsaicin onsensory C-type neurons and their axons in vitro. Neuroscience, 23,275-289.

OTSUKA, M. & YANAGISAWA, M. (1988). Effects of tachykinin antago-nists on a nociceptive reflex in the isolated spinal cord-tail prep-aration of the newborn rat. J. Physiol., 395, 255-270.

PETSCHE, U., FLEISCHER, E., LEMBECK, F. & HANDWERKER, H.O.(1983). The effects of capsaicin application to a pheripheral nerveon impulse conduction in functionally identified afferent nervefibres. Brain Res., 265, 233-240.

SZALLASI, A. & BLUMBERG, P.M. (1989). Resiniferatoxin, a phorbol-related diterpene, acts as an untrapotent analog of capsaicin, theirritant constituent of red pepper. Neuroscience, 30, 515-520.

SZALLASI, A. & BLUMBERG, P.M. (1990). Specific binding ofresiniferatoxin; an ultrapotent capsaicin analog by dorsal rootganglion membranes. Brain Res., 524, 106-111.

SZOLCSANYI, J. (1985). Sensory receptors and the antinociceptiveeffects of capsaicin. In Tachykinin Antagonists. ed. Hakenson, R. &Sundler, F. pp. 45-54. Amsterdam: Elsevier.

SZOLCSANYI, J. (1990). Capsaicin, irritation and desensitization:neurophysiological basis and future perspectives. In ChemicalSenses, Vol. 2, Irritation pp. 141-168. ed. Green, B.G., Mason, J.R.& Kare, M.R. New York: Marcel-Dekker.

SZOLCSANYI, J. & JANCSO-GABOR, A. (1975). Sensory effects of capsa-icin congenors. I. Relationship between chemical structure andpain producing potency of pungent agents. Arzneim. Forsch., DrugRes., 25, 1877-1881.

SZOLCSANYI, J. & JANCSO-GABOR, A. (1976). Sensory effects of capsa-icin congenors. II Importance of chemical structure and pungencyin desensitizing activity of capsaicin type compounds. Arzneim.Forsch., Drug Res., 26, 33-37.

THERIAULT, E., OTSUKA, M. & JESSELL, T. (1979). Capsaicin-evokedrelease of substance P from primary sensory neurons. Brain Res.,170, 209-213.

URBAN, L. & DRAY, A. (1990). Multiple actions of capsaicin on spinalsensory systems. Soc. Neurosci., 16, 231.

WINTER, J., DRAY, A., WOOD, J.N., YEATS, J.C. & BEVAN, S. (1990).Cellular mechanism of action of resiniferatoxin: a potent sensoryneuron excitotoxin. Brain Res., 520, 131-140.

YAKSH, T.L., FARB, D.H., LEEMAN, S.E. & JESSEL, T.M. (1979). Intra-thecal capsaicin depletes substance P in the rat spinal cord andproduces prolonged thermal analgesia. Science, 206, 481-483.

YANAGISAWA, M. & OTSUKA, M. (1984). The effect of a substance Pantagonist on chemically induced nociceptive reflex in the isolatedspinal cord-tail preparation of the newborn rat. Proc. Jpn. Acad.,60, Ser. B. 427-430.

(Received June 18, 1991Revised August 15, 1991

Accepted August 21, 1991)